According to a statistical study, the Milky Way contains a minimum of 100 billion planets.
On Jan. 11, 2012, one of our Hubble news releases trumpeted that our Milky Way Galaxy “Contains at Least 100 Billion Planets.” Since astronomers have detected less than 2,000 planets around other stars, that may seem like a huge extrapolation of the data. It would mean another 50 million planets for every one we have detected.
But wait, the math gets even more astonishing. Reading the full story, one finds that the conclusion is based upon a survey that examined only 40 stars. And in only three of those observations were planets detected. Can astronomers really go from three detections to the conclusion of 100 billion planets? Let’s examine the reasoning.
First, recognize that most of the planets detected so far are large planets (like Jupiter) and relatively close to their stars. The main planet-detection techniques are less sensitive for smaller and more distant planets. The survey that brought about this news release used a different technique, gravitational microlensing, which notes the change in brightness that occurs when a star and its orbiting planet pass in front of another star, distorting its light with their combined gravity.
Gravitational microlensing can better detect a variety of planet sizes, although the number of detections is very small. In addition, unlike other techniques that concentrate on specific types of stars, microlensing surveys the full range of stars and thus produces more general statistics.
As noted above, 40 microlensing events were analyzed for planets and three planets were detected. It is likely that many more planets existed in these systems, but were not detected. Each event was observed many times, but the signatures of planets are short-lived and could have occurred during the gaps in the data. Also, the geometry of a star-planet system could lead to a planet signal too weak or too short to be detected.
The estimation of this “detection efficiency” is a complex statistical analysis to quantify the ideas expressed in words above. Note that the analysis relied on previous results, increasing the total number of planet detections from three to 10. One has to consider a range of planet sizes, a range of orbit diameters, and a range of orbit inclinations. How many planets would there have to be in total, in order for the observations to find the detected number?
The results indicate that about 17% of the stars should have a Jupiter-sized planet, 52% should have a Neptune-sized planet, and 62% should have a “super-Earth” planet (a planet about 5-10 times the mass of Earth). Those percentages, as input to the statistical model, yield the observed number of planets from the survey.
The full conclusion comes when generalizing to all the stars of the Milky Way. The results indicate that, on average, each star has at least one planet. With about a hundred billion stars in our galaxy, that means at least 100 billion planets.
Finally, the question must be asked as to whether these results are robust. What is the level of uncertainty? The statistical analysis measures these uncertainties and, even being pessimistic, the survey results indicate at least 50 billion planets. On the other end of the spectrum, the statistics could justify as many as 200 billion planets. No matter where in this range the true number may lie, the result is both intriguing and exciting.
Billions and billions of planets — that’s a lot of places to explore.
